For a long time I've wondered why we humans usually describe colors as a 3 dimensional space (RGB) where it really only is a 1-dimensional one (wavelength). Took me sometime to realize this was because we have 3 types of rod cells ... and to start wondering how the world looks for the living creatures who have 4, 5 or more rod cells...
If, instead of measuring it, you asked naive subjects (probing for what they are conscious of), you might even get answers where red and violet are highly separated (as in a rainbow), but I think you can make them question that model easily by creating a palette going from red to violet that does not include orange, yellow, green, or blue.
In either case, the problem remains that humans perceive colors that do not physically exist, such as brown (aka 'dark orange-ish') and gray (aka 'shades of white')
Historically, some people would have known about mixing paints. That might have change their ideas about color, but that isn't RGB, either.
Finally, I am not sure one's color space will change with the addition of types of cone cells (it is cones, not rods that are color-sensitive. It alliterates: kones are for kolor).
maybe we would naturally understand fourier analysis in such a world? it's kind of hard to even think about.
maybe monochrome light would appear particularly wonderful (or pure) with such eyes? i was thinking about this the other day when i happened to be sitting in the path of some refracted light - it does appear particularly pure even though, with only three detectors, there's no real reason for it to do so (as far as i can see).